In a general image sensor, the charge stored in a light receiving unit (a photodiode) is read as a signal voltage, and is subjected to Analog/Digital (A/D) conversion (see Patent Document 1, for example).
According to the A/D conversion method disclosed in Patent Document 1, two A/D conversion circuits are connected to the same pixel output signal, and reference voltages Vref1 and Vref2 with different gradients are input from two reference voltage generating units to the respective A/D conversion circuits. In this manner, A/D conversion is performed with two gradation accuracies, so as to achieve a higher gradation accuracy and restrain increases in conversion time at the same time. In this case, however, the circuit area and the power consumption are doubled. Therefore, according to the method disclosed in Patent Document 1, only one A/D conversion circuit is used, and a determining unit is provided. The determining unit determines the size of each pixel output signal, and, in accordance with a result of the determination, selects one of the reference voltages Vref1 and Vref2 having two different gradients. In this manner, the use of different conversion accuracies depending on the sizes of pixel output signals is realized.